How to Use np.savetxt and np.loadtxt in NumPy

Use np.savetxt(filename, array) to save a NumPy array to a text file and np.loadtxt(filename) to load it back into an array. These functions handle plain text files and support options like delimiters and formatting for easy data exchange.

Syntax

np.savetxt saves an array to a text file with options for delimiter and formatting. np.loadtxt reads a text file and returns a NumPy array.

• np.savetxt(filename, array, delimiter=' ', fmt='%.18e'):
  filename - file path or object
  array - data to save
  delimiter - character separating values (default space)
  fmt - format string for numbers (default scientific)
• np.loadtxt(filename, delimiter=' ', dtype=float):
  filename - file path or object
  delimiter - character separating values
  dtype - data type of loaded array

python

np.savetxt(fname, X, delimiter=' ', fmt='%.18e')
np.loadtxt(fname, delimiter=' ', dtype=float)

Example

This example shows how to save a 2D array to a file and then load it back. It demonstrates using a comma as delimiter and formatting numbers with two decimals.

python

import numpy as np

# Create a 2D array
arr = np.array([[1.5, 2.3, 3.1], [4.7, 5.6, 6.2]])

# Save array to 'data.txt' with comma delimiter and 2 decimal places
np.savetxt('data.txt', arr, delimiter=',', fmt='%.2f')

# Load the array back from the file
loaded_arr = np.loadtxt('data.txt', delimiter=',')

print('Saved array:')
print(arr)
print('\nLoaded array:')
print(loaded_arr)

Output

Saved array:
[[1.5 2.3 3.1]
 [4.7 5.6 6.2]]

Loaded array:
[[1.5 2.3 3.1]
 [4.7 5.6 6.2]]

Common Pitfalls

Common mistakes include:

• Not matching the delimiter in np.loadtxt to the one used in np.savetxt, causing loading errors.
• Saving non-numeric data without specifying fmt, which can cause formatting issues.
• Trying to save/load arrays with mixed data types, which np.savetxt and np.loadtxt do not support well.

Example of a wrong and right way:

python

# Wrong: saving with comma delimiter but loading with default space delimiter
import numpy as np
arr = np.array([[1, 2], [3, 4]])
np.savetxt('wrong.txt', arr, delimiter=',')
try:
    np.loadtxt('wrong.txt')  # Will fail or load incorrectly
except Exception as e:
    print('Error:', e)

# Right: matching delimiter on load
loaded = np.loadtxt('wrong.txt', delimiter=',')
print('Loaded correctly:', loaded)

Output

Error: could not convert string to float: '1,2'
Loaded correctly: [[1. 2.]
 [3. 4.]]

Quick Reference

Function	Purpose	Key Parameters
np.savetxt	Save array to text file	filename, array, delimiter=' ', fmt='%.18e'
np.loadtxt	Load array from text file	filename, delimiter=' ', dtype=float

Key Takeaways

Always match the delimiter in np.loadtxt to the one used in np.savetxt.

Use the fmt parameter in np.savetxt to control number formatting.

np.savetxt and np.loadtxt work best with numeric arrays, not mixed types.

These functions save and load data in plain text files for easy sharing.

Check file paths and permissions to avoid file read/write errors.